Research suggests that the identification of subtypes of women drug users based on individual-level, dyad-level, and contextual variables may be essential to understanding unsafe sexual practices among drug-using women. This First Independent Research Support and Transition award research is designed to develop such subtypes, and explore a troubling pattern of risk behavior among Alaska Native women in particular. The specific aims of this project are: l. Develop a model to predict condom use by the sexual partners of drug using women, considering the following as potential predictors: power of the woman compared to her sexual partner, contextual variables related to sexual decision-making, psychosocial constructs, and selected demographic variables. In particular, investigate why White male - Alaska Native female sexual pairings have the lowest condom use of any sex/race combination in the Anchorage drug abuse cohort. 2. Determine whether homogeneous subgroups of women can be identified using variables from the condom use model. If so, describe the profiles of each subgroup. 3. Identify those behaviors, attitudes, beliefs, or circumstances that are believed by drug using women to be (a) predictive of condom use, and (b) modifiable. Document suggested strategies for change. 4. Delineate circumstances of sexual encounters in which condoms are versus are not used. The long-term goal of this research is to provide information that will be instrumental in the development of effective HIV-preventive interventions to increase levels of condom use among drug using women. A sample of women. confirmed by urinalysis to be current drug users, will be interviewed about their sexual behavior, drug use, and beliefs and experiences regarding a series of power- related, contextual, and psychosocial factors, and will also participate in one or more sets of focus groups and an individual in- depth interview.